Statistics and analyses of food safety inspection data and mining early warning information based on chemical hazards

Abstract

Food safety issues have grown in recent years, attracting great research attention. In addition to increasing supervision, regulatory authorities and related companies are currently trying to use the existing regulatory system to develop real-time data monitoring systems. In this study, food safety inspection data were collected from the information release platform of management departments across China. These data were processed with chemical hazardous substances as the objects of concern and were then classified according to hazardous substances, record product name, inspection location, and other categories. The current risk of chemical hazards in food was analyzed, and key information from the inspection data was mined. Frequent items and association rules of the inspection data were generated by the Apriori algorithm and evaluated according to the support, confidence, and rule interestingness (RI) to obtain key information to aid in developing an improved food safety inspection system. The results show that data mining methods can be used to obtain early warning information from food safety inspection data and are more efficient than traditional statistical methods. With data mining methods, an efficient early warning system can be established to assist management departments and manufacturers in ensuring food safety and quality.